U.S. patent application number 15/596920 was filed with the patent office on 2018-11-22 for adaptive trellis coding enabling and disabling in discrete multitone data communications.
The applicant listed for this patent is Adtran, Inc.. Invention is credited to Arlynn W. Wilson.
Application Number | 20180337807 15/596920 |
Document ID | / |
Family ID | 64176487 |
Filed Date | 2018-11-22 |
United States Patent
Application |
20180337807 |
Kind Code |
A1 |
Wilson; Arlynn W. |
November 22, 2018 |
ADAPTIVE TRELLIS CODING ENABLING AND DISABLING IN DISCRETE
MULTITONE DATA COMMUNICATIONS
Abstract
In a transmitter, first and second sets of discrete multitone
(DMT) sub-carrier signals or tones are identified. First and second
bit groups of a payload data frame corresponding to the first and
second sets of tones are selected. The first bit group is then
trellis encoded. The second bit group is not trellis encoded. The
first trellis coded tone group and the second bit group are then
constellation mapped to produce a DMT symbol for transmission. A
receiver may use an estimate of signal-to-noise ratio (SNR) of each
tone to determine whether to select the tone for inclusion in the
first or second set of tones. The receiver may provide the
transmitter with information indicating whether a tone is included
the first or second set of tones.
Inventors: |
Wilson; Arlynn W.;
(Huntsville, AL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Adtran, Inc. |
Huntsville |
AL |
US |
|
|
Family ID: |
64176487 |
Appl. No.: |
15/596920 |
Filed: |
May 16, 2017 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H03M 13/25 20130101;
H04M 3/007 20130101; H03M 13/353 20130101; H04L 5/006 20130101;
H03M 13/6533 20130101; H04L 1/0045 20130101; H04L 5/0007 20130101;
H04L 5/0046 20130101; H04L 47/6225 20130101; H04L 1/0003 20130101;
H03M 13/256 20130101; H04L 1/0041 20130101; H04L 25/03343 20130101;
H04L 1/006 20130101; H04B 3/32 20130101 |
International
Class: |
H04L 25/03 20060101
H04L025/03; H04L 5/00 20060101 H04L005/00; H04L 12/863 20060101
H04L012/863 |
Claims
1. A method for discrete multitone (DMT) data transmission,
comprising: identifying a first set of tones and a second set of
tones; selecting a first bit group comprising a first plurality of
bits of a payload data frame, the first bit group corresponding to
the first set of tones; selecting a second bit group comprising a
second plurality of bits of the payload frame, the second bit group
corresponding to the second set of tones; trellis encoding the
first plurality of bits to produce a first trellis coded tone
group; constellation mapping the first trellis coded tone group and
the second bit group to produce a DMT symbol; and transmitting the
DMT symbol via a communication channel.
2. The method of claim 1, wherein identifying a first set of tones
and a second set of tones comprises receiving a message indicating
whether a tone is included in one of the first set of tones and the
second set of tones.
3. The method of claim 2, wherein: the communication channel is one
of a very high-speed digital subscriber line (VDSL) channel and a
fast access to subscriber terminal (FAST) channel; and receiving a
message identifying a tone comprises receiving a message
identifying a tone via an Embedded Operations Channel.
4. The method of claim 2, wherein the message comprises a list of
one of the first set of tones and the second set of tones.
5. The method of claim 2, wherein the message indicates a
signal-to-noise ratio (SNR) of the tone identified by the message,
the method further comprising: comparing the SNR with a threshold;
and adding the tone identified by the message to one of the first
set of tones and the second set of tones in response to a result of
comparing the SNR with the threshold.
6. The method of claim 2, wherein the message is received by a
transmitter of the DMT symbol during a training operation preceding
a communication session between the transmitter and a receiver.
7. The method of claim 6, wherein the communication channel is one
of a very high-speed digital subscriber line (VDSL) channel and a
fast access to subscriber (FAST) channel.
8. The method of claim 2, wherein the message is received by the
transmitter during a communication session between the transmitter
and a receiver.
9. The method of claim 8, wherein the communication channel is one
of a very high-speed digital subscriber line (VDSL) channel and a
fast access to subscriber (FAST) channel.
10. A method for discrete multitone (DMT) data reception,
comprising: receiving a DMT symbol via a communication channel;
constellation de-mapping the DMT symbol to produce a first
plurality of bits defining a first bit group and a second plurality
of bits defining a second bit group, the first bit group
corresponding to a first set of tones, the second bit group
corresponding to a second set of tones; trellis decoding the first
plurality of bits to produce a plurality of trellis decoder output
bits; combining the trellis decoder output bits and the second
plurality of bits to produce a payload data frame; estimating a
signal-to-noise ratio (SNR) of each of a plurality of tones;
selecting the first set of tones and the second set of tones from
the plurality of tones by comparing the SNR of each of the
plurality of tones with a threshold; and transmitting a message via
a messaging channel identifying a tone whether a tone is included
in one of the first set of tones and the second set of tones.
11. The method of claim 10, wherein: the communication channel is
one of a very high-speed digital subscriber line (VDSL) channel and
a fast access to subscriber terminal (FAST) channel; and
transmitting a message identifying a tone comprises transmitting a
message identifying a tone via an Embedded Operations Channel.
12. The method of claim 10, wherein the message comprises a list of
one of the first set of tones and the second set of tones.
13. The method of claim 10, wherein the message is transmitted by a
receiver of the DMT symbol during a training operation preceding a
communication session between the transmitter and the receiver.
14. The method of claim 13, wherein the communication channel is
one of a very high-speed digital subscriber line (VDSL) channel and
a fast access to subscriber (FAST) channel.
15. The method of claim 10, wherein the message is transmitted by a
receiver of the DMT symbol during a communication session between
the transmitter and the receiver.
16. The method of claim 15, wherein the communication channel is
one of a very high-speed digital subscriber line (VDSL) channel and
a fast access to subscriber (FAST) channel.
17. A transmitter system for discrete multitone (DMT) data
communication, comprising: a memory configured to store information
indicating whether each of a plurality of tones is included in one
of a first set of tones and a second set of tones; a data frame
buffer configured to buffer a payload data frame; a bit extractor
configured to select a first bit group comprising a first plurality
of bits of the data frame buffer and a second bit group comprising
a second plurality of bits of the data frame buffer, the first bit
group corresponding to the first set of tones, the second bit group
corresponding to the second set of tones; a trellis encoder
configured to encode the first plurality of bits to produce a first
trellis coded tone group; and a constellation mapper configured to
produce a DMT symbol in response to the first trellis coded tone
group and the second bit group for transmission via a communication
channel.
18. The transmitter system of claim 17, further comprising a
messaging system configured to receive a message indicating whether
a tone is included in one of the first set of tones and the second
set of tones.
19. The transmitter system of claim 18, wherein the message
comprises a list of one of the first set of tones and the second
set of tones.
20. The transmitter system of claim 18, wherein the messaging
system is configured to receive the message during a training
operation preceding a communication session between the transmitter
system and a receiver system.
21. The transmitter system of claim 20, wherein the communication
channel is one of a very high-speed digital subscriber line (VDSL)
channel and a fast access to subscriber (FAST) channel.
22. The transmitter system of claim 18, wherein the messaging
system is configured to receive the message during a communication
session between the transmitter and a receiver.
23. The transmitter system of claim 22, wherein the communication
channel is one of a very high-speed digital subscriber line (VDSL)
channel and a fast access to subscriber (FAST) channel
24. A receiver system for discrete multitone (DMT) data
communication, comprising: a constellation de-mapper configured to
de-map a DMT symbol received via a communication channel to produce
a first plurality of bits defining a first bit group and a second
plurality of bits defining a second bit group, the first bit group
corresponding to a first set of tones, the second bit group
corresponding to a second set of tones; a trellis decoder
configured to trellis decode the first plurality of bits to produce
a plurality of trellis decoder output bits; a combiner configured
to combine the trellis decoder output bits and the second plurality
of bits to produce a payload data frame; a signal-to-noise ratio
(SNR) estimator configured to estimate a signal-to-noise ratio
(SNR) of each of a plurality of tones; a tone selector configured
to select the first set of tones and the second set of tones from
the plurality of tones by comparing the SNR of each of the
plurality of tones with a threshold; and a messaging system
configured to transmit a message via a messaging channel
identifying a tone whether a tone is included in one of the first
set of tones and the second set of tones.
25. The receiver system of claim 24, wherein: the communication
channel is one of a very high-speed digital subscriber line (VDSL)
channel and a fast access to subscriber terminal (FAST) channel;
and transmitting a message identifying a tone comprises
transmitting a message identifying a tone via an Embedded
Operations Channel.
26. The receiver system of claim 24, wherein the message comprises
a list of one of the first set of tones and the second set of
tones.
27. The receiver system of claim 24, wherein the message is
transmitted by a receiver of the DMT symbol during a training
operation preceding a communication session between the transmitter
and the receiver.
28. The receiver system of claim 27, wherein the communication
channel is one of a very high-speed digital subscriber line (VDSL)
channel and a fast access to subscriber (FAST) channel.
29. The receiver system of claim 24, wherein the message is
transmitted by a receiver of the DMT symbol during a communication
session between the transmitter and the receiver.
30. The receiver system of claim 29, wherein the communication
channel is one of a very high-speed digital subscriber line (VDSL)
channel and a fast access to subscriber (FAST) channel.
Description
BACKGROUND
[0001] Digital Subscriber Line (DSL) refers to data communication
technologies that leverage the longstanding infrastructure of
bundled twisted-pair copper wires that were originally deployed for
analog telephone service as a way of delivering high-speed Internet
access. Enhancements to the earliest DSL technology have led to
Asymmetric DSL (ADSL), Very High Bit-rate Digital Subscriber Line
(VDSL), and VDSL2, among others. The VDSL2 protocol is set forth in
the International Telecommunications Union (ITU) standard G.993.2.
VDSL2 employs Discrete Multitone (DMT) modulation, which is a form
of orthogonal frequency division multiplexing (OFDM). In DMT
modulation, the information-carrying frequency band associated with
a communication is divided into multiple (e.g., up to several
thousand) mutually orthogonal carrier frequencies, also referred to
as tones or sub-carrier frequencies. As the sub-carrier frequencies
into which a communication is divided are mutually orthogonal,
there is no interference between them. However, there may be
interference, i.e., crosstalk, between different communications
occurring on neighboring wire pairs in a cable bundle. There are
two types of crosstalk that can adversely affect a receiver:
Far-End Crosstalk (FEXT) and Near-End Crosstalk (NEXT). FEXT is
produced by neighboring wire pairs at a transmitter remote from the
affected receiver. NEXT is produced by neighboring wire pairs at a
transmitter local to the affected receiver.
[0002] Bit loading, a feature employed in VDSL transceivers,
adaptively allocates the number of bits that are transmitted on
each sub-carrier signal. The bit loading feature determines the
number of bits to allocate to a sub-carrier signal in response to a
measurement of signal-to-noise ratio (SNR) on the sub-carrier
signal. Thus, the bit loading feature allocates more bits (and
consequently, more bits per second per Hz) to a tone having a
higher measured SNR than to a tone having a lower measured SNR. A
common set of bit loading levels (i.e., numbers of bits that can be
adaptively allocated) consists of integer bits between 1 and 15. A
transceiver can adaptively select one of these bit loading levels
based on the current measured SNR of a tone.
[0003] VDSL2 transceivers also employ a feature known as vectoring,
which is a noise cancellation technique. Vectoring is used in VDSL2
specifically to cancel FEXT. Vectoring is described in ITU standard
G.993.5, "Self-FEXT cancellation (vectoring) for use with VDSL2
transceivers" (2010).
[0004] VDSL and VDSL2 transceivers employ trellis coding, which is
an error correcting technique. Trellis coding improves throughput
under low SNR conditions. In trellis coding, the bits are divided
between "payload bits" (also referred to as "user bits") that
encode the information or "payload" of the communication and
trellis bits that carry the error-correcting information. For
example, at a bit loading level of 15 bits, a trellis bit is
allocated every two tones, effectively resulting in 141/2 user bits
per tone plus 1/2 trellis bit per tone. In an instance in which
trellis coding results in improved throughput, such as in an
instance in which SNR is low, the sacrifice of some user bits for
trellis bits results in a net "coding gain." Under the ITU VDSL
standards, trellis coding is mandatory on all tones. Thus,
conventionally, trellis coding is employed on all tones during all
VDSL communications.
SUMMARY
[0005] Embodiments of the invention relate to selectively applying
trellis coding to a first set of tones and not applying trellis
coding to a second set of tones in discrete multitone (DMT) data
communication.
[0006] In one aspect, an exemplary method for DMT data transmission
includes identifying a first set of tones and a second set of
tones. The method further includes selecting a first bit group
comprising a first plurality of bits of a payload data frame, and
selecting a second bit group comprising a second plurality of bits
of the payload frame. The first bit group corresponds to the first
set of tones, and the second bit group corresponds to the second
set of tones. The method still further includes trellis encoding
the first plurality of bits to produce a first trellis coded tone
group. The second bit group, comprising the second plurality of
bits, is not trellis coded. The method also includes constellation
mapping the first trellis coded tone group and the second bit group
to produce a DMT symbol. The DMT symbol is transmitted via a
communication channel.
[0007] In another aspect, an exemplary method for DMT data
reception includes receiving a DMT symbol via a communication
channel, and constellation demapping the DMT symbol to produce a
first plurality of bits defining a first bit group and a second
plurality of bits defining a second bit group. The method further
includes trellis decoding the first plurality of bits to produce a
plurality of trellis decoder output bits. The method still further
includes combining the trellis decoder output bits and the second
plurality of bits to produce a payload data frame. The method also
includes estimating a signal-to-noise ratio (SNR) of each of a
plurality of tones, and selecting a first set of tones and a second
set of tones from the plurality of tones by comparing the SNR of
each of the tones with a threshold. Information indicating whether
a tone is included in the first set of tones or the second set of
tones may be sent to a transmitter via a messaging channel.
[0008] In still another aspect, an exemplary transmitter system for
DMT data communication includes a data frame buffer configured to
buffer a payload data frame, a memory configured to store
information indicating whether each tone is included in a first set
of tones or a second set of tones, and a bit extractor. The bit
extractor is configured to select from the data frame buffer first
and second bit groups corresponding to the first and second sets of
tones. The system further includes a trellis encoder configured to
encode the first bit group to produce a first trellis coded tone
group. The second bit group is not trellis coded. The system still
further includes a constellation mapper configured to produce a DMT
symbol in response to the first trellis coded tone group and the
second bit group for transmission via a communication channel.
[0009] In yet another aspect, an exemplary receiver system for DMT
data communication includes a constellation de-mapper configured to
de-map a DMT symbol received via a communication channel to produce
a first plurality of bits defining a first bit group and a second
plurality of bits defining a second bit group, and a trellis
decoder configured to decode the first plurality of bits to produce
a plurality of trellis decoder output bits. The receiver system
further includes a combiner configured to combine the trellis
decoder output bits and the second plurality of bits to produce a
payload data frame. The exemplary system may further include an SNR
estimator configured to estimate an SNR of each of a plurality of
tones. The receiver system may include a tone selector configured
to select a first set of tones and a second set of tones from the
plurality of tones by comparing the SNR of each of the plurality of
tones with a threshold. The receiver system may also include a
messaging system configured to transmit a message via a messaging
channel indicating whether a tone is included in first set of tones
or the second set of tones.
[0010] Other systems, methods, features, and advantages will be or
become apparent to one with skill in the art upon examination of
the following figures and detailed description. It is intended that
all such additional systems, methods, features, and advantages be
included within this description, be within the scope of the
specification, and be protected by the accompanying claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] The invention can be better understood with reference to the
following drawings. The components in the drawings are not
necessarily to scale, emphasis instead being placed upon clearly
illustrating the principles of the present invention.
[0012] FIG. 1 illustrates an exemplary system for DMT data
communication, in accordance with an exemplary embodiment of the
invention.
[0013] FIG. 2 illustrates a first set of tones and a second set of
tones.
[0014] FIG. 3 is an example of a plot illustrating a comparison of
various exemplary signals relating to exemplary methods for DMT
data communication.
[0015] FIG. 4 is another example of a plot similar to FIG. 3.
[0016] FIG. 5 is a block diagram illustrating a transmitter portion
of an exemplary system for DMT data communication, in accordance
with an exemplary embodiment of the invention.
[0017] FIG. 6 is a block diagram illustrating a receiver portion of
an exemplary system for DMT data communication, in accordance with
an exemplary embodiment of the invention.
[0018] FIG. 7 is a flow diagram illustrating an exemplary method
for DMT data communication, in accordance with an exemplary
embodiment of the invention.
[0019] FIG. 8A is a flow diagram illustrating another exemplary
method for DMT data communication, in accordance with an exemplary
embodiment of the invention.
[0020] FIG. 8B is a continuation of the flow diagram of FIG.
8A.
[0021] FIG. 9 is a block diagram of a transceiver, in accordance
with an exemplary embodiment of the invention.
DETAILED DESCRIPTION
[0022] As illustrated in FIG. 1, in an illustrative or exemplary
embodiment of the invention, a discrete multitone (DMT) data
communication system 100 includes a transmitter portion 102, a
receiver portion 104, and a communication link 106. The DMT data
communication system 100 may be a portion of, for example, a very
high bit-rate digital subscriber line (VDSL) system or,
alternatively, a fast access to subscriber terminal (FAST) system.
For example, transmitter portion 102 may be a digital subscriber
line access multiplexer (DSLAM) of a VDSL system, and receiver
portion 104 may be customer premises equipment (CPE) of a VDSL
system. Communication link 106 may include optical fiber, twisted
pair, or other media, configured to provide one or more
communication channels between transmitter portion 102 and receiver
portion 104. Although not shown for purposes of clarity,
transmitter portion 102 and receiver portion 104 each may be
included in a transceiver. Accordingly, communication via
communication link 106 may be bidirectional. As described in
further detail below, transmitter portion 102 may include, among
other elements, a trellis encoder 108, and receiver portion 104 may
include, among other elements, a trellis decoder 110. A messaging
link 112 is configured to communicate messages or control
information between receiver 104 and transmitter 102. Messaging
link 112 may, for example, comprise communications in accordance
with VDSL Embedded Operations Channel (EOC) standards. Although
messaging link 112 is shown for purposes of clarity as separate
from communication link 106, VDSL EOC messages may be transmitted
via the same physical medium as communication link 106.
[0023] As illustrated in FIG. 2, the systems and methods described
herein involve selecting, among all tones 202 (i.e., DMT
sub-carrier signals), a first set of tones 204 and a second set of
tones 206. First set of tones 204 and second set of tones 206 are
mutually exclusive, i.e., have no common tones 202. Although in the
example illustrated in FIG. 2, first set of tones 204 is higher in
frequency than second set of tones 206, the terms "first" and
"second" are used herein only for descriptive purposes to
distinguish one from the other, and should not be construed as
implying an order or other relationship. Thus, in other examples,
the DMT sub-carrier signals or tones of a "second" set may be
higher in frequency than those of a "first" set. The total number
of tones 202 (i.e., the sum of the number of tones 202 in first set
of tones 204 and second set of tones 206) may be any number. The
boundary or threshold frequency 208 between first set of tones 204
and second set may be any frequency. Although for purposes of
clarity the tones 202 in first set of tones 204 and second set of
tones 206 are depicted in FIG. 2 as contiguous, tones 202 may be
grouped in bands, spaced apart from one another in frequency (e.g.,
by guard bands), as well understood by one of ordinary skill in the
art. A reference herein to "all" tones 202 means all tones that are
on or active during a payload frame transmission. Although in the
example illustrated in FIG. 2 each set of tones consists of
contiguous tones, in other examples (not shown), portions of one
set of tones may alternate or otherwise be interspersed with
portions of the other set of tones, and the tones of each set
identified through an indexing array.
[0024] In an example illustrated in FIG. 3, a separation 302
between a received signal 304 and far-end crosstalk (FEXT) 306 on
communication link 106 (FIG. 1) represents a signal-to-noise ratio
(SNR). As may be appreciated by one of ordinary skill in the art,
received signal 304 and FEXT 306 depend upon factors such as the
length of the wires (commonly referred to as a "loop") in
communication link 106, the wire gauge, etc., and their plots in
FIG. 3 represent only one example of the effect of such
factors.
[0025] Transmitter portion 102 (FIG. 1) may employ bit loading
(also known as bit allocation). In a VDSL system, a common set of
bit loading levels (i.e., numbers of bits that can be adaptively
allocated) consists of an integer number of bits between 1 and 15.
Similarly, in a FAST system, a common set of bit loading levels
consists of an integer number of bits between 1 and 12. A
transceiver can adaptively select a bit loading level based on the
current measured SNR of a tone. The above-referenced separation 302
between received signal 304 and FEXT 306 represents an SNR that is
achieved in an example in which bit loading is not employed. In an
example in which bit loading is employed: the separation between
received signal 304 and FEXT 306 represents an SNR 308 that would
be required to provide a 3-bit loading level; the separation
between received signal 304 and FEXT 306 represents an SNR 310 that
would be required to provide a 6-bit loading level; the separation
between received signal 304 and FEXT 306 represents an SNR 312 that
would be required to provide a 9-bit loading level; the separation
between received signal 304 and FEXT 306 represents an SNR 314 that
would be required to provide a 12-bit loading level; and the
separation between received signal 304 and FEXT 306 represents an
SNR 316 that would be required to provide a 15-bit loading
level.
[0026] The noise level represented by FEXT 306 effectively limits
the number of bits that can be allocated to any tone 202 (FIG. 2).
Accordingly, in the example shown in FIG. 3, FEXT 306 limits
transmitter portion 102 to allocating maximums of: 15 bits to tones
at less than a frequency 318; 12 bits to tones at less than a
frequency 320; 9 bits to tones at less than a frequency 322; 6 bits
to tones at less than a frequency 324; and less than 6 bits to the
remaining tones at greater than frequency 324.
[0027] Transmitter portion 102 may employ trellis coding. Trellis
coding may achieve a net coding gain even though trellis coding
sacrifices some payload bits by transmitting trellis code bits
during the time in which payload bits would otherwise be
transmitted. The net amount of coding gain achieved by trellis
coding in a particular instance of operation depends upon how many
tones are trellis coded, how many bits are allocated to each of
those tones by bit loading, and the amount of Reed Solomon coding
and interleaver concatenated as an outer Forward Error Correction.
Conventionally, in accordance with ITU VDSL standards G.992.3/5,
G.993.2 and G.993.5, multidimensional WEI trellis coding is
mandatory on all tones. Generally, trellis coding provides a
greater coding gain when SNR is low and fewer bits are allocated to
more tones than when SNR is high and more bits are allocated to
fewer tones, because coding gain decreases as more payload bits are
utilized. In the example shown in FIG. 3, in the presence of FEXT
306, essentially no coding gain is likely achieved by trellis
coding tones allocated 15 bits, i.e., tones at less than frequency
318, because bit loading limits allocation of 15 bits to only the
relatively few tones at less than frequency 318 (e.g., less than a
few hundred kHz).
[0028] Transmitter portion 102 may employ vectoring to reduce or
eliminate FEXT 306. For example, if transmitter portion 102 were to
reduce FEXT 306 to FEXT 306' by employing vectoring or some other
technique, transmitter portion 102 could allocate a greater number
of bits to a greater number of tones. In the presence of FEXT 306'
instead of FEXT 306, transmitter portion 102 could allocate 15 bits
to tones at less than a frequency 326. However, trellis coding
tones at less than frequency 326 may result in little net coding
gain, no coding gain, or even a negative net coding gain because
the negative effect of sacrificing the relatively large number of
payload bits (of the relatively large number of tones allocated 15
bits) weighs heavily against the positive effect of trellis coding.
Note that if transmitter portion 102 were to essentially eliminate
FEXT 306 entirely, transmitter portion 102 could allocate 15 bits
to all tones.
[0029] The example illustrated in FIG. 4 is similar to the example
illustrated in FIG. 3, but illustrates a common effect of a longer
loop (i.e., communication link 106). More specifically, the
received signal 404 and FEXT 406 are more attenuated at higher
frequencies compared with received signal 304 (FIG. 3). Transmitter
portion 102 (FIG. 1) may utilize any integer bit loading level
between, for example, 1 and 15. In the example illustrated in FIG.
4: the separation between received signal 404 and FEXT 406
represents an SNR 408 that would be required to provide a 3-bit
loading level; the separation between received signal 404 and FEXT
406 represents an SNR 410 that would be required to provide a 6-bit
loading level; the separation between received signal 404 and FEXT
406 represents an SNR 412 that would be required to provide a 9-bit
loading level; the separation between received signal 404 and FEXT
406 represents an SNR 414 that would be required to provide a
12-bit loading level; and the separation between received signal
404 and FEXT 406 represents an SNR 416 that would be required to
provide a 15-bit loading level. In this example, FEXT 406
effectively imposes a limit that prevents allocating 15 bits to
essentially any tones.
[0030] Transmitter portion 102 may employ vectoring to reduce or
eliminate FEXT 406. For example, if transmitter portion 102 were to
reduce FEXT 406 to FEXT 406' by employing vectoring or some other
technique, transmitter portion 102 could allocate a greater number
of bits to a greater number of tones. In the presence of FEXT 406'
instead of FEXT 406, transmitter portion 102 could allocate 15 bits
to tones at less than a frequency 420. However, trellis coding
tones at less than frequency 420 may result in little net coding
gain, no coding gain, or even a negative net coding gain because
the negative effect of sacrificing the relatively large number of
payload bits (of the relatively large number of tones allocated 15
bits) weighs heavily against the positive effect of trellis
coding.
[0031] In the example shown in FIG. 4, even if transmitter portion
102 were to employ vectoring or other technique to eliminate FEXT
406 entirely, a noise floor 422 (e.g., at -140 dBm/Hz) would
effectively limit transmitter portion 102 in the manner graphically
represented by the intersections of bit loading levels 408-416 with
noise floor 422. That is, transmitter portion 102 would be limited
to allocating a maximum of: 15 bits to tones at less than a
frequency 424; 12 bits to tones at less than a frequency 426; 9
bits to tones at less than a frequency 428; 6 bits to tones at less
than a frequency 430; and 3 bits to the remaining tones at greater
than frequency 430. Trellis coding tones at less than frequency 424
is likely to result in a negative net coding gain because the
negative effect of sacrificing the relatively large number of
payload bits (of the relatively large number of tones allocated 15
bits) weighs heavily against the positive effect of trellis
coding.
[0032] The above-described examples illustrate that enabling
trellis coding of a first set of tones while disabling trellis
coding of a second set of tones may avoid a negative net coding
gain. In the example described above with regard to FIG. 4, in
which all FEXT is eliminated and only noise floor 422 limits bit
loading, the systems and methods described herein may selectively
enable trellis coding of a first set of tones at greater than
frequency 424 and selectively disable trellis coding of a second
set of tones at less than frequency 424. As described below with
regard to exemplary methods, in a given instance of operation, a
boundary or threshold frequency, such as frequency 424 in the
foregoing example, may be related to SNRs of the tones and bit
loading levels.
[0033] As illustrated in FIG. 5, a transmitter 500 may be an
example of transmitter portion 102 (FIG. 1). Transmitter 500 may
be, for example, a near-end transmitter in a VDSL system or,
alternatively, a near-end transmitter in a FAST system. Transmitter
500 includes a data frame buffer 502, a trellis encoder 504, a
constellation mapper 506, a parallel-to-serial converter 508, a
messaging system 510, a memory 512 configured to store a tone list
514 or other information indicating whether each tone is included
in the first set of tones or the second set of tones, and a bit
extractor 516. In addition to the foregoing elements, other
elements commonly included in conventional VDSL or FAST
transmitters as known to one of ordinary skill in the art may be
included in transmitter 500 but are not shown for purposes of
clarity.
[0034] Data frame buffer 502 is configured to store one or more
payload data frames, i.e., the data to be transmitted. Bit
extractor 516 is configured to extract a plurality of (L) bits from
data frame buffer 502. Bit extractor 516 is further configured to
select, from among the L bits, a first bit group 518 comprising a
first plurality of (L-M) bits in data frame buffer 502 and a second
bit group 520 comprising a second plurality of (M) bits in data
frame buffer 502. The first and second bit groups 518 and 520
correspond to the first and second sets of tones, respectively.
Accordingly, tone list 514 serves as a control input to bit
extractor 516, and bit extractor 516 extracts or selects first bit
group 518 and second bit group 520 based on tone list 514.
[0035] The first bit group 518 is provided directly to the input of
trellis encoder 504, which is configured to encode first bit group
518 to produce a first trellis coded tone group 522. First trellis
coded tone group 522 is provided to a portion of the input of
constellation mapper 506. Second bit group 520 is provided directly
to another portion of the input of constellation mapper 506,
bypassing trellis encoder 504. Constellation mapper 506 is
configured to map the combination of first trellis coded tone group
522 and second (non-trellis coded) bit group 520 to Quadrature
Amplitude Modulation (QAM) points. Using an inverse fast-Fourier
transform (IFFT), parallel-to-serial converter 508 is configured to
convert the parallel-format data that is output by constellation
mapper 506 to a serial data format to produce a DMT symbol.
Transmitter 500 may transmit a stream of such DMT symbols via
communication link 106 (FIG. 1).
[0036] Messaging system 510 is configured to receive a message from
a remote source, such as a receiver 600 described below with regard
to FIG. 6. As well understood by one of ordinary skill in the art,
conventional VSDL systems may employ a messaging system that allows
transmitters and receivers to exchange messages using messaging
channel 112 (FIG. 1). The message may, for example, comprise tone
list 514 or elements thereof, or other information identifying one
or more tones and indicating whether a tone is included in the
first set of tones or the second set of tones. Alternatively, as
described in further detail below, the message may comprise
information identifying one or more tones and information enabling
transmitter 500 to determine whether to include a tone in the first
set of tones or the second set of tones. Messaging system 510 may
store or otherwise update tone list 514 in memory 512 in response
to receiving the message. Other aspects of the messaging system may
be conventional and are therefore not described herein.
[0037] Tone list 514 may be organized in any manner For example,
tone list 514 may include information individually identifying each
tone in the first set and the second set. Alternatively, tone list
514 may include information identifying each tone in only one of
the first and second sets but not the other. Alternatively, tone
list 514 may include information identifying a range of tones, such
as a lowest tone and a highest tone, in one or both of the first
and second sets. For example, the received message may include such
information identifying a range of tones, and messaging system 510
may use the information to update tone list 514.
[0038] As illustrated in FIG. 6, a receiver 600 may be an example
of receiver portion 104 (FIG. 1). Receiver 600 includes a DMT
symbol sampler 602, a serial-to-parallel converter 604, a
constellation de-mapper 606, a trellis decoder 608, an SNR
estimator 610, a tone list generator 612, a messaging system 614,
and a memory 616 configured to store a tone list 618. In addition
to the foregoing elements, other elements commonly included in
conventional VDSL or FAST receivers as known to one of ordinary
skill in the art may be included in receiver 600 but are not shown
for purposes of clarity.
[0039] Receiver 600 receives DMT symbols. For example, receiver 600
may be a far-end VDSL or FAST receiver that receives a stream of
DMT symbols from transmitter 500 (FIG. 5) via communication link
106 (FIG. 1). DMT symbol sampler 602 is configured to provide each
received DMT symbol to serial-to-parallel converter 604.
Serial-to-parallel converter 604 is configured to convert the
received DMT symbol from a serial data format to a parallel data
format using a fast Fourier transform (FFT). The parallel-format
data that is output by serial-to-parallel converter 604 is provided
to the input of constellation de-mapper 606. Constellation
de-mapper 606 is configured to de-map the parallel-format data from
QAM points.
[0040] The output of constellation de-mapper 606 is provided to a
data splitter 620. Based on tone list 618, data splitter 620
selects, from among the resulting data bits that define the output
of constellation de-mapper 606, a first set of resulting data bits
622 corresponding to the first set of (trellis coded) tones and a
second set of resulting data bits 624 corresponding to the second
set of (non-trellis coded) tones. The first set of resulting data
bits 622 is provided to the input of trellis decoder 608. Trellis
decoder 608 is configured to decode the first set of resulting data
bits 622 into L-M decoded bits. A data combiner 626 combines the
L-M decoded bits with the second set of M resulting data bits 624
to define a payload data frame, i.e., the received data.
[0041] The outputs of constellation de-mapper 606 and trellis
decoder 608 are also provided to SNR estimator 610, which is
configured to estimate the SNR of each tone. Tone list generator
612 is configured to select the first and second sets of tones by
comparing the SNR of each tone with a threshold. Tone list
generator 612 may generate tone list 618 or a portion thereof, or
information otherwise indicating whether a tone is included in the
first set of tones or the second set of tones, as described above
with regard to FIG. 5.
[0042] Messaging system 614 may be configured to transmit a message
to transmitter 500 (FIG. 5) via messaging link 112 comprising tone
list 618 or elements thereof (e.g., a tone or range of tones for
transmitter 500 to add to remove from its tone list 514).
Transmitter 500 may store or otherwise maintain tone list 514 based
on the messaging information that transmitter 500 receives from
receiver 600. Alternatively, or in addition, receiver 600 may
receive information from transmitter 500 via messaging link 112
that aids receiver 600 in compiling or otherwise maintaining tone
list 618. For example, receiver 600 could send a message to
transmitter 500 identifying one or more tones and associated SNR
values. Transmitter 500 could compare the SNR values with a
threshold, update its tone list 514 in accordance with the result
of the comparison, and share the result with receiver 600 via
messaging link 112. Through the messaging described herein, tone
lists 514 and 618 may be maintained essentially identical to each
other during normal communication of payload data.
[0043] As illustrated in FIG. 7, an exemplary method 700 for DMT
data communication may be performed by a device such as, for
example, above-described transmitter 500 (FIG. 5). In the following
descriptions of exemplary methods, although certain acts or steps
described below naturally precede others for the exemplary
embodiments to operate as described, the invention is not limited
to the order of those acts or steps if such order or sequence does
not alter the functionality of the invention. That is, it is
recognized that some acts or steps may be performed before, after,
or in parallel (i.e., substantially simultaneously) with other acts
or steps without departing from the scope and spirit of the
invention. In some instances, certain acts or steps may be omitted
or not performed, without departing from the scope and spirit of
the invention. Further, words such as "thereafter," "then," "next,"
etc., are not intended to limit the order of the acts or steps.
Rather, such words are used as aids in guiding the reader through
the descriptions of the exemplary methods.
[0044] As indicated by block 702, first and second sets of tones
(i.e., DMT sub-carrier signals) are identified. For example,
transmitter 500 may access above-described tone list 514. Tone list
514 or elements thereof may be received from another device, such
as, for example, receiver 600 (FIG. 6) via messaging link 112.
Alternatively, or in addition, transmitter 500 may update tone list
514 in response to receiving a message from another device, such
as, for example, receiver 600 via messaging link 112. Transmitter
500 may receive or update tone list 514 at any time using and may
use coordinated signaling. For example, tone list 514 may be
received or updated during a training operation preceding a
communication session between transmitter 500 and receiver 600.
Alternatively, or in addition, tone list 514 may be received or
updated during such a communication session.
[0045] As indicated by block 704, first and second bit groups, each
comprising a plurality of bits of a payload data frame, are
selected. The first and second bit groups correspond to the first
and second sets of tones. As indicated by block 706, the first bit
group is trellis coded. The second bit group is not trellis coded.
As indicated by block 708, the first (trellis coded) bit group and
the second (non-trellis coded tone group) are constellation mapped
to QAM complex values, using an IFFT, to produce a time domain DMT
symbol corresponding to the DMT payload data. As indicated by block
710, the DMT symbols are transmitted via a channel (e.g.,
communication link 106).
[0046] As illustrated in FIGS. 8A-8B, an exemplary method 800 for
DMT data communication may be performed by a device such as, for
example, above-described receiver 600 (FIG. 6). As indicated by
block 802 (FIG. 8A), a DMT symbol is received via a channel (e.g.,
communication link 106) from a transmitter, such as transmitter 500
(FIG. 5). Note that the set of tones that transmitter 500 used to
produce the DMT symbol consists of the first and second sets of
tones. As indicated by block 803, the DMT symbol is converted from
serial to parallel (e.g., using a FFT). As indicated by block 804,
the received DMT symbol is constellation de-mapped. The first set
of resulting data bits are then trellis decoded, as indicated by
block 806. A second set of resulting data bits are combined with
the first set to produce the resulting payload data frame 808 or
otherwise used in a conventional manner. For example, the payload
data frame 808 may be processed using Reed Solomon Decoding.
[0047] As indicated by block 810, SNRs of all tones are estimated.
The SNR of a tone may be estimated using the inputs and outputs of
the constellation demapping and trellis decoding described above
with regard to blocks 804 and 806, respectively. The SNR may be
estimated in a conventional manner, such as by the method commonly
referred to as constellation slicing or decision. In constellation
slicing, the differences between the complex value before and
following the slice define the error. The magnitude of the error
represents the noise for use in SNR estimation.
[0048] As indicated by block 812, the SNR of a tone is compared
with a threshold. If it is determined (block 812) that the SNR of a
tone exceeds the threshold, and the bit loading level is a maximum,
then the tone is added to or otherwise included in the second set
of tones, as indicated by block 814. As described above, the second
set of tones will not be trellis coded. If it is determined (block
812) that the SNR of a tone does not exceed the threshold, or if
the bit loading level is not the maximum, then the tone is added to
or otherwise included in the first set of tones, as indicated by
block 816. As described above, the first set of tones will be
trellis coded.
[0049] The "maximum bit loading level," as the term is used herein,
is the bit loading level that carries the greatest weight in
whether the net coding gain is negative or not negative (i.e., zero
or positive). In the example described above with regard to FIG. 4,
the 15-bit loading level weighs more heavily in this respect than
the other bit loading levels. In that example, trellis coding tones
below frequency 420 would undesirably result in a net negative net
coding gain. Accordingly, tones below frequency 420 would be added
to or otherwise included in the second set of tones, whereas tones
above frequency 420 would be added to or otherwise included in the
first set of tones. In this manner, the first and second sets of
tones are selected. As indicated by block 818, the steps described
above with regard to blocks 810-816 are applied to all tones. When
the steps described above with regard to blocks 810-816 have been
applied to all tones, tone list 514 (FIG. 5) may be generated or
updated.
[0050] It should be understood that references in this disclosure
to specific bit loading levels or maximum bit loading levels, such
as 15 bits or 12 bits, is intended only to be exemplary. For
example, although maximums of 15 bits and 12 bits are commonly
employed in VDSL and FAST systems, respectively, other bit loading
levels may be employed.
[0051] Tone list 514 (FIG. 5) may be maintained by transmitter 500
based on information generated by receiver 600 and transmitted to
transmitter 500 via messaging link 112. For example, receiver 600
may generate tone list 618 by estimating SNR in the manner shown in
FIGS. 5 and 8A, in which VDSL receiver 500 compares each SNR value
with a threshold and adds the tone to the first or second set in
accordance with the result of the comparison. Alternatively, in
other embodiments receiver 600 could send the SNR values per tone
to transmitter 500 via messaging link 112, and transmitter 500
could compare the SNR values with a threshold and add the tone to
the first or second set (e.g., in a tone list) in accordance with
the result of the comparison. In such an embodiment, transmitter
500 could share the resulting tone list with receiver 600 or advise
receiver 600 of the tone or tones to be added to the tone list by
sending a message via messaging link 112.
[0052] Generating or updating tone lists 514 and 618 or related
information in the manner described above may occur at any time.
For example, tone lists 514 and 618 may be generated, shared,
updated, etc., during a training operation preceding a
communication session between transmitter 500 and receiver 600.
Alternatively, or in addition, tone lists 514 and 618 may be
generated, shared, updated, etc., during such a communication
session. As indicated by block 820 (FIG. 8B), it may be determined
whether a condition has occurred. The condition may include whether
a threshold amount of time has elapsed since previously generating,
sharing, updating, etc., one or both of tone lists 514 and 618.
Alternatively, or in addition, the condition may include whether a
tone that has been added to or otherwise included in the second set
of tones (block 814) was previously included in the first set of
tones, or whether a tone that has been added to or otherwise
included in the first set of tones (block 816) was previously
included in the second set of tones. If it is determined (block
820) that the condition has occurred, then a message is transmitted
from receiver 600 to transmitter 500, as indicated by block 822.
This message may include tone list 618 or, alternatively, may
include update information that transmitter 500 can use to update
tone list 514. Such update information may include information
identifying one or more tones for transmitter 500 to add to or
remove from tone list 514. Alternatively, such update information
may include information identifying one or more tones for
transmitter 500 to move within tone list 514, so as to indicate a
change in association with the first and second sets of tones.
[0053] A tone list may be represented in any form. For example, a
tone list may comprise a series of ones and zeros corresponding to
the list of tones currently in use, where the ones represent
membership in the first set of (trellis coded) tones and zeroes
represent membership in the second set of (non-trellis coded)
tones, or vice versa.
[0054] Method 800 may be performed repeatedly during a
communication session, as a stream of DMT symbols is received. It
should be understood that for purposes of clarity the method steps
are described above without regard to relative timing. For example,
portions of method 800, such as the portion indicated by blocks
812-822, may be performed less frequently than other portions of
method 800, such as the portion indicated by blocks 802-808.
[0055] As illustrated in FIG. 9, a transceiver 900, which may be a
DSLAM, CPE, or other VDSL device or a FAST device, may include a
processing system 902 comprising a processor 904 and a memory 906.
Transceiver 900 may further include an Ethernet physical interface
908 and a VDSL physical interface 910. The above-described payload
data frames may be communicated via Ethernet physical interface 908
in a conventional manner as known to one of ordinary skill in the
art. The above-described stream of DMT symbols may he communicated
via VDSL physical interface 910 in a conventional manner as known
to one of ordinary skill in the art. Additional elements commonly
included in a conventional DSLAM, CPE, or other VDSL, or FAST
device as known to one of ordinary skill in the art may be included
in transceiver 900 but are not shown for purposes of clarity.
Processor 904 and memory 906 may communicate signals with each
other and with other devices via one or more interconnects 912,
such as buses. In accordance with conventional computing
principles, processor 904 operates under the control of software or
firmware code, which configures processing system 902 to control
various functions or methods, including the methods described
herein. Such methods may also include conventional methods
controlled by a DSLAM, CPE, or other VDSL device.
[0056] Transceiver 900 is configured with processing logic that can
include data frame buffer logic 914, trellis encoder logic 916,
constellation mapper logic 918, bit extractor logic 920, DTM symbol
sampler logic 922, constellation de-mapper logic 924, trellis
decoder logic 926, SNR estimator logic 928, tone list generator
logic 930, and messaging system logic 932. Also, although not shown
for purposes of clarity, tone list 514 (FIG. 5) and other data may
be stored in memory 906, which may be an example of memory 512
(FIG. 5).
[0057] Although data frame buffer logic 914, trellis encoder logic
916, constellation mapper logic 918, bit extractor logic 920, DTM
symbol sampler logic 922, constellation de-mapper logic 924,
trellis decoder logic 926, SNR estimator logic 928, tone list
generator logic 930, and messaging system logic 932 are shown in
FIG. 9 in a conceptual manner as stored in or residing in memory
906, one of ordinary skill in the art understands that such logic
elements arise through the operation of processor 904 in accordance
with conventional computing device principles. That is, software or
firmware contributes to programming or configuring processing
system 902 to be characterized by such logic elements. Although
memory 906 is depicted in FIG. 9 as a single or unitary element for
purposes of clarity, memory 906 can be of any suitable type and can
have any suitable structure, such as one or more modules, chips,
etc. Memory 906 can be of a non-volatile type, such as flash
memory. Likewise, although processor 904 is depicted in FIG. 9 as a
single or unitary element for purposes of clarity, processor 904
can be of any suitable type and can have any suitable structure,
such as one or more modules, chips, etc. For example, processor 904
can comprise one or more microprocessors or microcontrollers. Some
or all of the foregoing processing system elements can be provided
in, for example, an application-specific integrated circuit (ASIC)
or other integrated digital device. It should be understood that
the combination of memory 906 and the above-referenced logic
elements or software, firmware, instructions, etc., underlying the
logic elements, as stored in memory 906 in non-transitory
computer-readable form, defines a "computer program product" as
that term is understood in the patent lexicon. In view of the
descriptions herein, persons skilled in the art will readily be
capable of providing suitable software or firmware or otherwise
configuring switch 12 to operate in the manner described.
[0058] Data frame buffer logic 914 may contribute to the functions
described above with regard to, for example, block 702 (FIG. 7).
Messaging system logic 932 also may contribute to the functions
described above with regard to block 702. Also, processor system
902, as configured with data frame buffer logic 914, may serve as a
means for performing such functions. Bit extractor logic 920 may
contribute to the functions described above with regard to, for
example, block 704 (FIG. 7). Also, processor system 902, as
configured with bit extractor logic 920, may serve as a means for
performing such functions. Trellis encoder logic 916 may contribute
to the functions described above with regard to, for example, block
706 (FIG. 7). Also, processor system 902, as configured with
trellis encoder logic 916, may serve as a means for performing such
functions. Constellation mapper logic 918 may contribute to the
functions described above with regard to, for example, block 708
(FIG. 7). Also, processor system 902, as configured with
constellation mapper logic 918, may serve as a means for performing
such functions. Also, processor system 902, as configured with
constellation mapper logic 918, may serve as a means for performing
such functions. DMT symbol sampler logic 922 may contribute to the
functions described above with regard to, for example, block 802
(FIG. 8A). Also, processor system 902, as configured with DMT
symbol sampler logic 922, may serve as a means for performing such
functions. Constellation de-mapper logic 924 may contribute to the
functions described above with regard to, for example, block 804
(FIG. 8A). Also, processor system 902, as configured with
constellation de-mapper logic 924, may serve as a means for
performing such functions. Trellis decoder logic 926 may contribute
to the functions described above with regard to, for example, block
806 (FIG. 8A). Also, processor system 902, as configured with
trellis decoder logic 926, may serve as a means for performing such
functions. SNR estimator logic 928 may contribute to the functions
described above with regard to, for example, block 810 (FIG. 8A).
Also, processor system 902, as configured with SNR estimator logic
928, may serve as a means for performing such functions. Tone list
generator logic 930 may contribute to the functions described above
with regard to, for example, blocks 812-818 (FIG. 8A). Also,
processor system 902, as configured with tone list generator logic
830, may serve as a means for performing such functions. Messaging
system logic 932 may contribute to the functions described above
with regard to, for example, blocks 820-822 (FIG. 8B). Also,
processor system 902, as configured with messaging system logic
932, may serve as a means for performing such functions.
[0059] One or more illustrative or exemplary embodiments of the
invention have been described above. However, it is to be
understood that the invention is defined by the appended claims and
is not limited to the specific embodiments described.
* * * * *